|
|
|
|
|
|
by rs186
310 days ago
|
|
|
> GPT-5 stood out for its analytical strength and review clarity. The sentence is too obviously LLM generated, but whatever. > Weaknesses: > > False positives: A few reviews include incorrect or harmful fixes. > Inconsistent labeling: Occasionally misclassifies the severity of findings or touches forbidden lines. > Redundancy: Some repetition or trivial suggestions that dilute review utility. wtf are "forbidden lines"? |
|
|